---

name: retro preamble-tier: 2 version: 2.0.0 description: | Weekly engineering retrospective. Analyzes commit history, work patterns, code quality, contributors, persistent history, and trends. Use when asked to "weekly retro", "what did we ship", or "engineering retrospective". Proactively suggest at the end of a work week or sprint. (gstack) allowed-tools:

• Bash
• Read
• Write
• Glob
• AskUserQuestion triggers:
• weekly retro
• what did we ship
• engineering retrospective gbrain: schema: 1 context_queries:
  • id: prior-retros kind: filesystem glob: "~/.gstack/projects/{repo_slug}/retros/*.md" sort: mtime_desc limit: 5 render_as: "## Prior retros for this project"
  • id: recent-timeline kind: filesystem glob: "~/.gstack/projects/{repo_slug}/timeline.jsonl" tail: 30 render_as: "## Recent timeline events"
  • id: recent-learnings kind: filesystem glob: "~/.gstack/projects/{repo_slug}/learnings.jsonl" tail: 10 render_as: "## Recent learnings"

---

{{PREAMBLE}}

{{BASE_BRANCH_DETECT}}

/retro — Weekly Engineering Retrospective

Generate an engineering retrospective from commit history, work patterns, quality metrics, and team contributions. Identify the caller, analyze every contributor, and produce specific praise plus growth opportunities for a senior IC/CTO-level builder using Claude Code as a force multiplier.

Arguments

• /retro — default: last 7 days
• /retro 24h — last 24 hours
• /retro 14d — last 14 days
• /retro 30d — last 30 days
• /retro compare — compare current window vs prior same-length window
• /retro compare 14d — compare with explicit window
• /retro global — cross-project retro across all AI coding tools (7d default)
• /retro global 14d — cross-project retro with explicit window

{{GBRAIN_CONTEXT_LOAD}}

Instructions

Parse the argument to determine the time window; default to 7 days. Report all times in the user's local timezone and do not set TZ.

Midnight-aligned windows: For d and w, compute an absolute local-midnight start date and use --since="YYYY-MM-DDT00:00:00" in git queries. Example: on 2026-03-18 with 7d, start at 2026-03-11T00:00:00. For w, multiply by 7 days. For h, use --since="N hours ago".

Argument validation: If the argument doesn't match a number followed by d, h, or w, the word compare (optionally followed by a window), or the word global (optionally followed by a window), show this usage and stop:

Usage: /retro [window | compare | global]
  /retro              — last 7 days (default)
  /retro 24h          — last 24 hours
  /retro 14d          — last 14 days
  /retro 30d          — last 30 days
  /retro compare      — compare this period vs prior period
  /retro compare 14d  — compare with explicit window
  /retro global       — cross-project retro across all AI tools (7d default)
  /retro global 14d   — cross-project retro with explicit window

If the first argument is global, skip repo mode and follow Global Retrospective Mode. Optional second argument is the window; default 7d. This mode does not require a git repo.

{{LEARNINGS_SEARCH}}

Standard Repo Retrospective Mode

{{RETRO_REPO_MODE}}

{{LEARNINGS_LOG}}

{{GBRAIN_SAVE_RESULTS}}

Global Retrospective Mode

{{RETRO_GLOBAL_MODE}}

---

Compare Mode

When the user runs /retro compare (or /retro compare 14d):

1. Compute current-window metrics using the same midnight-aligned start date as normal mode.
2. Compute the prior same-length window using midnight-aligned --since and --until to avoid overlap. Example: current 7d starting 2026-03-11T00:00:00 means prior is --since="2026-03-04T00:00:00" --until="2026-03-11T00:00:00".
3. Show a side-by-side comparison table with deltas and arrows
4. Write a brief narrative highlighting the biggest improvements and regressions
5. Save only the current-window snapshot to .context/retros/ (same as a normal retro run); do not persist the prior-window metrics.

Tone

• Encouraging but candid; specific to actual commits/code.
• Praise must be earned, concrete, and 1:1-ready. Growth suggestions should read as worthwhile investment advice, not criticism.
• Never compare teammates negatively; each person's section stands alone.
• Target 3000-4500 words, slightly longer for teams.
• Use markdown tables/code blocks for data and prose for narrative.
• Output to conversation only, except the .context/retros/ JSON snapshot.

Important Rules

• Narrative output goes to the conversation. The only repo-mode file write is .context/retros/ JSON.
• Use origin/<default> for git queries, not stale local main.
• Use local timestamps; do not override TZ.
• If the window has zero commits, say so and suggest another window.
• Round LOC/hour to nearest 50 and treat merge commits as PR boundaries.
• Do not read CLAUDE.md or other docs; this skill is self-contained.
• On first run, skip comparison sections gracefully.
• Global mode needs no git repo, writes snapshots to ~/.gstack/retros/, skips missing AI tools, compares only same-window global retros, and displays a 365-day streak cap as 365+ days.